EXHIBIT 101 / THE HANDBOOK
EXHIBIT#92HIGH
Technology & POSComposite score: 53/100

AI Model Manipulation

A technically sophisticated actor feeds an AI fraud-detection system false data specifically to blind it.

Risk Scores

Frequency20
Financial Loss80
Detection Difficulty90
Prevention Ease70
Composite Score53/100

Perpetrator Roles

IT / Tech

Affected Formats

QSRFast CasualCloud Kitchen

Detection Difficulty

Extremely difficult to detect without automated monitoring. Manual audits rarely catch this scheme.

HOW IT WORKS
Staff meal policy in place
Meals prepared beyond entitlement
Excess meals consumed or removed
Food cost rises
Variance attributed to waste
Handbook Chapter · F-092

Staff Meal Policy Abuse

Full chapter with identity card, real examples, warning signs, detection and prevention guidance.

Exhibit AI Detection

AI-Powered Detection Design

Monitor detection-model performance and flag rate over time, specifically watching for a change that doesn't correspond to any real operational shift.

Prevention Guidance

Audit Controls

Standard audit procedures are effective. Regular cash counts and reconciliation catch this scheme early.

Technology

Automated POS monitoring, real-time exception reporting, and AI-driven anomaly detection significantly reduce exposure.

Culture & Training

Staff awareness programs, anonymous reporting channels, and clear consequence policies deter opportunistic fraud.

Seen this in your operation?

Submit an anonymous report to help build the next edition of the handbook.

Report This Scheme